Hi, you mentioned in your paper "To provide the exemplar for our method, we first train a plain translation network to generate natural images and use them to retrieve the exemplars from the dataset.". Can you provide the pretrained translation network? I want to generate exemplar for another face dataset. Thanks!